Akamai Technologies Inc

NASDAQ:AKAM

Good afternoon, and welcome to the Akamai Technologies, Inc. Fourth Quarter 2023 Earnings Conference Call. [Operator Instructions]. Please note this event is being recorded. I would now like to turn the conference over to Tom Barth, Head of Investor Relations. Please go ahead.

Thank you, operator. Good afternoon, everyone, and thank you for joining Akamai's Fourth Quarter 2023 Earnings Call. Speaking today will be Tom Leighton, Akamai's Chief Executive Officer; and Ed McGowan, Akamai's Chief Financial Officer. Please note that today's comments include forward-looking statements, including statements regarding revenue and earnings guidance. These forward-looking statements are subject to risks and uncertainties and involve a number of factors that could cause actual results to differ materially from those expressed or implied by such statements. The factors include any impact from macroeconomic trends, the integration of any acquisitions and any impact from geopolitical developments. Additional information concerning these factors is contained in Akamai's filings with the SEC, including our annual report on Form 10-K and quarterly reports on Form 10-Q. The forward-looking statements included in this call represent the company's view on February 13, 2024. Akamai disclaims any obligation to update these statements to reflect new information, future events or circumstances, except as required by law. As a reminder, we will be referring to some non-GAAP financial metrics during today's call. A detailed reconciliation of GAAP and non-GAAP metrics can be found under the financial portion of the Investor Relations section of akamai.com. Before I turn the call over to Tom, I'd like to let everyone know that I have transitioned ahead of Akamai Investor Relations role to Mark Stoutenberg. Many of you have met Mark over the past year, and I'm confident you will find him extremely helpful about all things Akamai. I'll be moving into another role internally, and I want to thank Tom personally for bringing me into the Akamai culture 10 years ago. It's been a privilege to work with so many wonderful people both here at Akamai, and of course, all of you externally. I wish you all happiness and good fortune. And now I'd like to turn the call over to Tom.

Thanks, Tom, and thank you very much for the terrific job that you've done over the last 10 years at Akamai. It's truly been a pleasure working with you. And I'd like to join you in welcoming Mark as your successor. Turning now to our Q4 results. I'm pleased to report that Akamai delivered a strong finish to a very successful 2023. Fourth quarter revenue grew to $995 million, and non-GAAP operating margin was 30%. Non-GAAP earnings per share in Q4 was $1.69 up 23% year-over-year and up 22% in constant currency. For the full year, revenue was $3.81 billion, and non-GAAP operating margin was 30%. Full year non-GAAP earnings per share was $6.20, up 15% year-over-year and up 16% in constant currency. Security revenue in Q4 grew 17% year-over-year and was up 18% in constant currency, contributing nearly half of our total revenue in the quarter. Security growth was driven in part by continued strong demand for our market-leading Guardicore segmentation solution as more enterprises relied on Akamai to defend against malware and ransomware. Customers who purchased segmentation in Q4 included one of the largest global tech firms in India and a leading payment systems company based in Hong Kong that handles 15 million transactions a day. We've also seen strong interest in our new API security solution that we announced in August. Early adopters of this new solution include three major financial institutions, several major retail brands in Europe and North America and a leading industrial automation company in Europe. Our customers are seeing the value of this new capability with several already paying over $0.5 million per year for the service. Akamai's API security solution also earned recognition from industry analysts in Q4. KuppingerCole named Akamai an overall industry leader in their API security and management Leadership Compass Report. And Gartner validated our strong and expanded API capabilities in its market guide for cloud, web application and API protection. Turning now to cloud computing. I'm pleased to say that we accomplished what we set out to achieve last year in terms of infrastructure deployment, product development, jump-starting our partner ecosystem, onboarding the first mission-critical apps from some major enterprise customers and achieving substantial cost savings as we moved our own applications from hyperscalers to the Akamai Connected Cloud. After launching Akamai Connected Cloud last year, we rolled out 14 new core computing regions around the world, giving us a total of 25 overall. We also enhanced our cloud compute offering by doubling the capacity of our object storage solution and adding premium instances for large commercial workloads that are designed to deliver consistent performance with predictable resource and cost allocation. Also, our Akamai global load balancer is now live. This integrated service is designed to route traffic request to the optimal data center to minimize latency and ensure no single point of failure. Last year, we also amplified our go-to-market approach with our cloud computing partner program. The collaborative nature of the program provides a unique model for Akamai to engage with customers at a consultative level to deeply understand their requirements and pain points and to provide a complete solution, leveraging the combined strength of the partner's technology and Akamai's distributed compute platform. We've already partnered with several leading SaaS and PaaS providers in cloud, data and processing platforms. We're very pleased with the progress that we've made thus far and are looking forward to adding more new partnerships in 2024. We've also begun to gain traction with some of our largest customers as they migrate mission-critical apps on to our cloud platform. For example, one of the world's best known social media platforms spent approximately $5 million on computing services with us last year, and they're already on a run rate to do more than double that this year. And two of the world's best-known software companies recently signed on and are already spending about $0.25 million per month on cloud computing with Akamai. In Q4, we also signed up a major global media measurement and analytics company, and we displaced a hyperscaler when we signed BluTV, the largest VOD streaming company in Turkey. The streaming customer told us that they found our platform to be simple to use, automated and intuitive cloud agnostic for a smooth multi-cloud migration and affordable with very low egress costs, all backed by a trusted and reliable partner. In addition to exceeding our full year cloud computing revenue goal of $500 million last year, we also derive significant cost savings by migrating several of our own applications from hyperscalers to Akamai Connected Cloud. Our Bot Manager and Enterprise Application Access Solutions were among the first to migrate. Together, these products are used by over 1,000 customers and they generate over $300 million in annual revenue for Akamai. But all that is just the beginning. Today, we're excited to announce the next phase of our multiyear strategy to transform the cloud marketplace, taking cloud computing to the edge by embedding cloud computing capabilities into Akamai's massively distributed edge network. Akamai's new initiative, codenamed Gecko, which stands for Generalized Edge Compute combines the computing power of our cloud platform with the proximity and efficiency of the edge to put workloads closer to users than any other cloud provider. Traditional cloud providers support virtual machines and containers in a relatively small number of core data centers. Gecko is designed to extend this capability to our edge [ pops ] bringing full stack computing power to hundreds of previously hard-to-reach locations. Deploying our cloud computing capabilities into Akamai's worldwide Edge platform will also enable us to take advantage of existing operational tools, processes and observability, enabling developers to innovate across the entire continuum of compute and providing a consistent experience from centralized cloud to distributed edge. Nobody else in the marketplace does this today. In the latest implementation phase that we're announcing today, Akamai aims to embed compute with support for virtual machines into about 100 cities by the end of the year. We've deployed new Gecko architected regions in four countries already as well as in cities that lack a concentrated hyperscaler presence. These initial locations are listed in today's press release. Following that, we plan to add support for containers. And then we plan to develop automated workload orchestration to make it easier for developers to build applications across hundreds of distributed locations. We've been conducting early trials of Gecko with several of our enterprise customers that are eager to deliver better experiences for their customers by running workloads closer to users, devices and sources of data. Their early feedback has been very encouraging as they evaluate Gecko for tasks such as AI inferencing, deep learning for recommendation engines, data analytics, multiplayer gaming, accelerating banking transactions personalization for e-commerce and a variety of media workflow applications such as transcoding. In short, I'm incredibly excited for the prospects of Gecko as we move full stack compute to the edge. Turning now to content delivery. I'm pleased to report that Akamai remains the market leader by a wide margin, providing the scale and performance required by the world's top brands as we help them deliver reliable, secure and near-flawless online experiences. Our delivery business continues to be an important generator of profit that we use to develop new products to fuel Akamai's future growth. And it's an important driver of our security and cloud computing businesses as we harvest the competitive and cost advantages of offering delivery, security and compute on the same platform as a bundle. As we've noted in past calls, we're selective when it comes to less profitable delivery opportunities. And this is a discipline that we intend to maintain and in some cases, increase in 2024. To be clear, we still aim to be the best in delivery for our customers, and we believe that our disciplined approach will benefit our business by allowing us to focus more of our investment in security and cloud computing which are now approaching 2/3 of Akamai's revenue and growing at a rapid rate. In summary, we're pleased to have accomplished what we said we would do in 2023. Our cloud computing plans are taking shape as we envisioned. Our expanded security portfolio is enabling us to deepen our relationships with customers, and we continue to invest in Akamai's future growth, while also enhancing our profitability. Now I'll turn the call over to Ed for more on our results and our outlook for Q1 and 2024. Ed?

Thank you, Tom. I would also like to thank Tom Barth for his incredible service for 10 years as our Head of Investor Relations. Now moving on to our results. Let me start by saying that I'm very pleased with our fiscal 2023 year results, delivering $6.20 of non-GAAP earnings per share, capping off a year of double-digit earnings growth for our shareholders. Today, I'll cover our Q4 results, provide some color regarding 2024, including some items to help investors better understand a few factors that will impact our upcoming results, and enclosed with our Q1 and full year 2024 guidance. Starting with revenue. Q4 revenue was $995 million, up 7% year-over-year as reported and in constant currency. We saw continued strong growth in our compute and security businesses during the fourth quarter. Our compute business grew to $135 million, up 20% year-over-year as reported and in constant currency. We continue to be very pleased with the feedback regarding our cloud compute offerings, and we are very optimistic about the early traction we are seeing from enterprise customers. Moving to security. Revenue was $471 million, growing 18% year-over-year and up 17% in constant currency. Our security revenue continues to be driven by strong growth in our Guardicore segmentation solution and our industry-leading web app firewall, denial of service and bought management solutions. In addition, and as Tom mentioned, we are encouraged by the early traction of our new API security solution. During the fourth quarter, we signed 17 API security customers, including 4 with annual contract values in excess of $500,000 per year. Our delivery revenue was $389 million, including approximately $20 million from the contracts we recently acquired from StackPath and Lumen. International revenue was $479 million, up 8% year-over-year or up 6% in constant currency representing 48% of total revenue in Q4. Finally, foreign exchange fluctuations had a negative impact on revenue of $4 million on a sequential basis and a positive $6 million benefit on a year-over-year basis. Moving to profitability. In Q4, we generated strong non-GAAP net income of $263 million or $1.69 of earnings per diluted share up 23% year-over-year or 22% in constant currency and $0.07 above the high end of our guidance range. These stronger-than-expected EPS results were driven primarily by continued progress on the cost-saving initiatives we have previously outlined and approximately $6 million in lower-than-expected transition services or TSA costs associated with the StackPath and Lumen contracts as our services organization migrated the customers onto our platform much faster than we expected. Q4 CapEx was $143 million or just below 15% of revenue. We were very pleased with our continued focus on lowering the capital intensity of our delivery business. This effort, along with our very strong profitability enabled us to deliver very strong free cash flow results in Q4. Moving to our capital allocation strategy. During the fourth quarter, we spent approximately $55 million to buy back approximately 500,000 shares. For the full year, we spent approximately $654 million to buy back approximately 8 million shares. We ended 2023 with approximately $500 million remaining on our current repurchase authorization. Going forward, our intention is to continue buying back shares to offset dilution from employee equity programs over time and to be opportunistic in both M&A and share repurchases. Before I move on to guidance, there are several items that I want to highlight in order to give investors some greater insight into the business. The first relates to our delivery revenue. In the first half of 2024, 7 of our top 10 CDN customers' contracts come up for renewal, as we've discussed in the past, this type of renewal generally leads to an initial drop in revenue. And then we typically see revenue grow again as traffic increases over time. We have factored the expected outcome of these renewals into our Q1 and full year 2024 guidance. In addition, as Tom mentioned, we plan to continue to optimize our delivery business by focusing on how we charge certain high-volume traffic customers for their usage on our network, all with an eye on profitability. For example, we plan to start charging a premium for higher cost delivery destinations. We expect to continue to optimize the ratio of peak to day-to-day traffic and we plan to negotiate different pricing for API traffic versus download traffic. choosing to shed some less profitable traffic will result in a more balanced and profitable approach to pricing, which we believe is the right strategy for the company. Second, OECD member countries continue to work toward the enactment of a 15% global minimum corporate tax rate. And in particular, in December 2023, the Swiss Federal Council declared the rules in effect for Switzerland beginning in 2024. As a result, we anticipate that our non-GAAP effective tax rate will increase by roughly 1.5 to 2 percentage points to approximately 18.5% to 19%. We estimate this increase in our tax rate will have a negative impact on Q1 non-GAAP EPS of approximately $0.02 to $0.03 per diluted share and a negative impact on full year non-GAAP EPS of approximately $0.12 to $0.15 per diluted share. The impact of this tax rate change has been factored into our Q1 and full year 2024 guidance. Third, we expect to generate significantly more free cash flow in 2024 compared to 2023 levels. This is primarily due to much lower capital expenditures in 2024, along with our continued focus on profitability. Please note that the full cost to build out our Gecko compute sites we announced earlier today is included in our Q1 and full year 2024 capital expenditure guidance. Fourth, I want to remind you of the typical seasonality we experienced on the top and bottom lines throughout the year. Regarding revenue, the fourth quarter is usually our strongest quarter. Regarding profitability. In Q1, we incurred much higher payroll taxes related to the reset of social security taxes for employees who maxed out during 2023 and from stock vesting from employee equity programs, which tend to be more heavily concentrated in the first quarter. It's also worth noting that in Q3, our annual company-wide merit-based salary increases go into effect so we tend to see higher operating costs in Q3 compared to Q2 levels. Finally, for 2024, we anticipate heightened volatility in foreign currency markets driven by the unpredictable timing and magnitude of Federal Reserve policy changes and their impact on interest rates. With this in mind, forecasting the trajectory of FX in the latter part of the year poses a formidable challenge. Thus, for the full year, we plan to provide annual revenue growth, security and compute revenue growth, non-GAAP EPS growth, non-GAAP operating margin and CapEx only in constant currency, based on [ 12/31/2023 ] exchange rates. However, for the coming quarter, we will provide both as reported and constant currency guidance. As a reminder, we have approximately $1.2 billion of annual revenue that is generated from foreign currency with the euro, yen, great British pound being the largest non-U.S. dollar sources of revenue. In addition, our costs in non-U.S. dollars tend to be significantly lower than our revenue and are primarily in Indian rupee Israeli Shekel, and Polish zloty. Moving now to guidance. Our guidance for 2024 assumes no material changes, good or bad in the current macroeconomic landscape. For the first quarter of 2024, we are projecting revenue in the range of $980 million to $1 billion or up 7% to 9% as reported or 8% to 10% in constant currency over Q1 2023. At current spot rates, foreign exchange fluctuations are expected to have a positive $2 million impact on Q1 revenue compared to Q4 levels and a negative $4 million impact year-over-year. At these revenue levels, we expect cash gross margin of approximately 73% as reported and in constant currency. Q1 non-GAAP operating expenses are projected to be $305 million to $310 million. We anticipate Q1 EBITDA margins of approximately 42% to 43% as reported and in constant currency. We expect non-GAAP depreciation expense of $127 million to $129 million. We expect non-GAAP operating margin of approximately 29% to 30% as reported and in constant currency for Q1. And with the overall revenue and spend configuration I just outlined, we expect Q1 non-GAAP EPS in the range of $1.59 to $1.64 or up 14% to 18% as reported and 16% to 19% in constant currency. This EPS guidance assumes taxes of $56 million to $58 million based on an estimated quarterly non-GAAP tax rate of approximately 18.5% to 19%. It also reflects a fully diluted share count of approximately 155 million shares. Moving on to CapEx. We expect to spend approximately $146 million to $154 million, excluding equity compensation and capitalized interest in the first quarter, this represents approximately 15% of total revenue. Looking ahead to the full year for 2024, we expect revenue growth of 6% to 8% in constant currency. We expect security revenue growth of approximately 14% to 16% in constant currency we expect compute revenue growth to be approximately 20% in constant currency, and we are estimating non-GAAP operating margin of approximately 30% in constant currency. And full year CapEx is expected to be approximately 15% of total constant currency revenue, which again includes the Gecko compute build-outs. We expect our CapEx to be roughly broken down as follows: approximately 3% of revenue for our delivery and security business, approximately 4% of revenue for compute, approximately 7% of revenue for capitalized software and the remainder for IT and facility-related spend. We expect non-GAAP earnings per diluted share growth of 7% to 11% in constant currency. And as we mentioned earlier, this non-GAAP earnings guidance is based on a non-GAAP effective tax rate of approximately 18.5% to 19% and a fully diluted share count of approximately 155 million shares. In closing, we are very pleased with the strong finish to 2023. We continue to be excited about our growth prospects and driving profitability across the business. Now Tom and I would like to take your questions. Operator?

[Operator Instructions] Our first question will come from Fatima Boolani of Citi.

Ed, I wanted to drill into the delivery, the segment performance and the guidance and some modeling points that you shared with us. I was hoping you could parse out for us some of the organic traffic trends you saw in the platform parse away from the traffic from the acquired contracts between StackPath and Lumen. And then to the extent you can talk about in [ time slam ] you have with regards to absorbing some of this additional acquired traffic from these contracts. And the last piece of the question here is also the guidance we calculated down 6% for delivery in 2024, implied by your guidance, full year guidance. Can you talk to what proportion of that traffic but you would have deemed lower quality being peeled off is maybe the reason why we're not seeing a more sharper improvement in the delivery franchise? And then I have a quick follow-up.

All right, sure. Let me see if I can tackle all those for you. I'll start with the performance in Q4. So I'll break it down into a couple of buckets. So in Q4, we tend to see stronger seasonal quarter for us. And we did see some of that. So from the retail customer perspective, we saw bursting roughly to the tune of about half of what we saw last year. And I think that has to do mostly with the zero overage as that's become more popular, we are seeing less bursting. If I look back a few years, that's down about 50% -- '22 compared to '21 and again, '23 compared to '22. Some of it could be macro factors, hard to tell. We're not really the best gauge for folk's consumption, but more of their surfing. If I look at gaming, gaming was a better year. If I look last year, gaming was pretty weak this year. It was pretty good. Gaming tends to be fairly lower-priced delivery, so you don't get as much upside in revenue. Video was up quarter-over-quarter, but not as much as we saw in the prior year. So I would say kind of an okay fourth quarter from delivery. The one area that was probably the weakest was more in sort of high tech. So think about that as new connected devices, maybe it's connected TV or a new, say, an iPad or something like that, printer drivers, that sort of stuff. That was much weaker than what we saw last year. Last year was a pretty good bump up in that category. So that really sort of sums up what the dynamics are in terms of the delivery business in Q4. And I looked at that not including the acquisition traffic. The second question you asked, if I wrote it down correctly, was the absorption of the traffic from the acquisitions. We've largely done that. The services team did a phenomenal job migrating over the customers. We did it in probably about 1/3 of the time that we had expected to be able to do that. There's still a few stragglers, but transition services costs are immaterial. So that's why we didn't call that out. But that's largely been done. In terms of this year, why are we seeing a down 6%, if that's the math you're using for delivery. A lot of it has to do with the renewals. Unfortunately, these larger customers, as we've talked in the past, we only have about 8 customers that are greater than 1% of revenue. And unfortunately, 7 of them are renewing in the first half of the year. That does have a bit of an impact on the delivery business. The majority of the revenue from those customers comes from delivery. We've seen this in the past. It seems like every two years, we kind of bump into this. Even though we try to sign staggered contracts what happens often is someone might sign a 2- or a 3-year deal with a revenue commit, they'll spend that in a shorter period of time, and we just get a combination of renewals here. And then the other thing was on the shedding of traffic. So let me just get a little bit more specific with what we're doing. So last year, if you remember, we talked about being a bit more stringent with the peak to average ratios. We're going to continue to do that. We did a pretty good job. There's still a few customers we need to adjust. And that generally happens where customers will split, and they have lots of day-to-day traffic and they might split that among three or four CDMs and then they'll have peak events, whether that's a big video event or a big download event, and we get the disproportionate size of that peak. So you want to make sure you're compensated for that. So the way to do that is you get a higher percentage of the day-to-day or you just limit the peak. So we'll continue doing that. And sometimes, that may mean we lose a little bit of traffic, but that's okay because you can see the results in CapEx. The new thing that we're doing now is we've worked with product and IT to enable us to be able to build for very granular level on destinations. In the past, we generally would build based on large geos, say, for example, the U.S. or Europe or Asia, now we're able to get one level deeper. And so we can go after some of the areas where our costs are a bit higher to deliver. It's unclear. We're assuming that customers have choice. They may decide to take that traffic somewhere else, some of our competitors don't really focus as much on profit as we do. If that's the case, then that's okay with us. So I think as you -- as we try to factor all that in, those are the two driving factors why we're not seeing a better delivery performance in '24. I think I hit all your questions.

And then easy one, just on the gross margins, you gave us the guidance for cash gross margin in the first quarter. But just qualitatively, if you can talk to us about some of the puts and takes at a full year level as we think about the mix of business changing your focus on pricing discipline and just even U.S. international mix, any sort of puts and takes in terms of the dynamics we should be thinking about COGS and cash gross margins for the full year? And that's it for me.

Sure. No problem. In terms of the cost of goods sold, I'll start with the good stuff that's going on in cost of goods sold. Obviously, you touched on the mix. So as we get a higher degree of security business, that obviously helps our gross margin line. And the second thing is the movement of our third-party cloud costs onto our own platform. We did a great job so far this year. Tom talked a bit about that in his prepared remarks. And we have more to go, and we have a road map to get that. So that will drive some significant savings. The downside or what's sort of [ trapping ] them or not expanding the margins as much, I should say, we could be going a bit higher is that with the build out of the Gecko sites, we'll incur some additional co-location costs. So as we start to build out those facilities, you don't have revenue to go against it. But also, as I talked about last year, when we built out some of the bigger sites, with the accounting rules, when you do long-term [ Colo ] commitments or any kind of commitment for long-term leases, you have to straight line any commitments. So we end up with a bit of a disconnect between what we're paying in cash and what we're actually expensing. So we're able to offset that with the savings we get from the third-party cloud and the mix, but those are the underlying dynamics.

The next question comes from Frank Louthan of Raymond James.

How different is Gecko from what you're doing now with compute? Is this a new packaging of some of your products? Or is it something a little different? And then secondly, of the CDN business that you acquired last year, how much are you expecting that to fall off? I think the goal was to try to get it from other services. But what are you factoring in there?

Yes, I'll take the first question. Yes, Gecko is not packaging. Gecko is a new capability where we're going to be offering full stack compute in 100 cities by the end of the year and ultimately, in hundreds of cities. We're actually taking full stack compute the kinds of services you get in the node or a hyperscaler and making that available in our edge pops. Now today, the edge pops, we have 4,000 and they run function as a service. We'll spin up JavaScript apps and 4,000 locations in over 700 cities in milliseconds based on user demand. But that's not full-stack compute. Now we're going to be taking virtual machines and containers and supporting those in our edge platform. And that enables our customers to get much better performance and scalability, also lower cost because of the financial benefits we get from our Edge platform. So it really becomes, I think, a very compelling cloud computing offering that just doesn't exist in the marketplace today. It's not a packaging thing at all. This is a new capability. And of course, ultimately, after we get the support for [ VMs ] and containers, we want to make it work just as we do function as a service, so that we'll be spinning up containers and VMs on demand, where and when they're needed. And that capability doesn't exist today in the market. Ed, do you want to take the CDN question, the acquisition?

Sure. Happy to do that. Frank, Yes. I would say, just as a reminder, when we acquired the businesses, we actually acquired selected customers. What we mean by that is we actually went through and we left some customers that we weren't going to take, for example, if someone violated our acceptable use policy, really small customers. And then believe it or not, some that had pricing that we just didn't want to take. So we had already gone through sort of a selection process. If you recall, I had given guidance last quarter of about $18 million to $20 million for this quarter, and we hit the high end of that range, which I'm happy that we did. So we've largely been able to migrate over everything that we had hoped to. There are a few customers that churn. But by and large, we've gotten everything out of that acquisition or those acquisitions, I should say, there are two of them that we had hoped to.

The next question comes from James Fish of Piper Sandler.

I wanted to circle over to security actually. Look, security was a little bit lower than I think we were all anticipating for Q4 and get that we have some drivers underneath that are helping the business. But did you see any push out of deals and maybe that's contributing to your confidence around mid-teens growth for 2024. Help us kind of on the confidence for sustained mid-teens growth. And really, how is that selling outside the installed base and penetration of those new security packages, [ gone ]?

Yes, I was actually pretty pleased with our performance, and I didn't see many deals push out. We didn't have any large license deals like we did in Q3. So that might be -- that always sort of skewed some of the results. So we don't have any of that. But no deals that really pushed out from a security perspective. Very pleased with what we're seeing with Guardicore, continued great growth there. That's been phenomenal. And that's -- a lot of customers are being driven new verticals and things like that. The channel has been joined phenomenally well there. We talked a lot about in our prepared remarks, API security. I don't get surprised often, but I've been surprised with the ARPUs there. I've been very pleased with that. That's been very, very good to see. And the strength in -- we talked about the bundles that we have done a lot on the last call. That's continued to go very well, and we're seeing very strong growth in our WAF, in our fraud products, too, bot management and fraud protector. So really strength across the board. Nothing so far from macroeconomic challenges that always can change, you'll never know what can happen. But nothing that we saw in Q4. As far as the projections for this year, we feel pretty confident. We've generally been relatively conservative with our approach to security growth. We don't factor in any major type of attacks where sometimes we'll see a spike in demand or anything like that. So we feel pretty good with how security is going.

Maybe just to click down one level into Fatima's earlier question on the CDN side and something that Tom said as well in his prepared remarks. Is there any way to understand how much on the CDN side of the base you plan to essentially -- I don't want to say give away for free, but give away for free to get that compute revenue or help us understand kind of the dynamic between what we should expect between CDN and compute kind of dollar shifting?

Yes. I think you can't factor in any percentage there at this point. We have been in some discussions with some very large media companies where we would offer discounted or free delivery and return for a significant portion of the compute business. On balance, that would be a great trade for us, much more profitable and much more revenue because at the end of the day, big media companies will spend 10x on compute, what they'll spend on delivery and even security. So this is something that we see the hyperscalers do. They sometimes give away the delivery return for getting the compute business because that's where the vast majority of the revenue is and very profitable. So we'll keep you advised as we go if that starts to make a material difference.

The next question comes from Keith Weiss of Morgan Stanley.

Just one for Ed. Throughout 2020, we talked a lot about [indiscernible] initiatives. We talked a lot about migrate [indiscernible] savings. And I guess just to put a kind of disappointing about this point about the lack of margin expansion in the [indiscernible]. Is there something holding that back? Or is there some incremental investments perhaps behind Gecko that we're making instead of that? Or a question like, why not more margin [indiscernible] given all the efficiency sort of improvements that you guys have been putting in [indiscernible] the [ past years ].

Yes, I'll take this one, Tom. You broke up a little bit there, Keith, but I think I got the genesis of the question. So yes, we've done -- I thought a great job of making some acquisitions over the last few years, investing in our compute business, spending an awful lot to build out our compute facilities, adding a lot of functionality, growing our security business, doing acquisitions there as well and got back to 30% margin last year and continuing it this year. We've always said that's been a pretty healthy spot to run the business. We're investing because we see opportunity for growth. And there's always a balance. You can't cut your way to greatness. Perhaps we could cut a few more points, but then what are we leaving on the table. I think we've been pretty disciplined and balanced with our approach in terms of investing for growth and returning to margin. We've got some pretty exciting areas. We're seeing great growth in API security. It's still early days there. The product will continue to get better. We're seeing good ARPUs there so I'm very excited about what we're doing there. We made investments in go-to-market in Guardicore, and that certainly has paid off and the investments in computer side show some early returns. So again, it's a balanced approach. We're in this for the long term, and I think we don't want to shoot ourselves in the foot and not go after some of these big opportunities that we have in front of us.

The next question comes from Mark Murphy of JPMorgan.

So Tom, you've done a very solid job with the compute business. And in the prepared comments, you mentioned onboarding of some mission-critical apps. Wondering if you could shed a little light on the pipeline of that kind of critical app that you see coming to you in 2024? And thus far, how well is your infrastructure handling the intensity of those larger workloads in terms of stability, reliability, uptime, et cetera? And then I have a quick follow-up.

Yes. So signing on compute customers is a big focus for us this year. We have the basic infrastructure in place. Of course, now we're building out Gecko. But just with the basic infrastructure already in 25 cities, we'll be looking to add on many more mission-critical apps from major enterprises. And some examples, you look at social media, live transcoding, we now have giant companies using that on the platform, one for live sports broadcasting, another for live user-generated content. Another customer, we're hosting e-commerce sites for them in a way that performance better because closer to the end user and less expensive. AI inferencing for ad targeting, personalizing content. And again, you want to do that really fast. And you just don't have time to backhaul that up into a centralized location. You want to be in a lot of locations around the world to get better performance. And again, we can do it at a lower cost. We've even got a large -- one of the world's largest banks now using us our edge compute to register credit cards, they'll use your credit cards with Apple Pay because Apple Pay requires you do the registration in 60 milliseconds. And the only way they can get that done fast is to do it on Akamai Connected Cloud. So really, a lot of different applications already on the platform, doing proofs of concept now. So the focus this year. And I think it's a good pipeline is to be taking on many more mission-critical apps for major enterprises. And of course, we're the first big one. We've got enormous applications already running on the platform and very successfully. And we do it in a multi-cloud way. And as I talked about earlier, now we have the global load balancing built in and the [ failover ] capability. so that it does make for a reliable service that's high performing. So really excited about what's coming this year.

Ed, I wanted to ask you, you're providing the fiscal year guidance in constant currency. And of course, we all understand it's going to be very difficult to predict the actual fluctuations in the spot market. But if we looked at it at current FX levels, do you think you would skew that a 6% to 8% revenue growth level higher or lower? I mean if we were to try to translate it into reported U.S. dollars in our models right now today?

Yes. Good question. And I think the CPI report today gave you a good view of how things can change so quickly. The market originally had thought there'd be a lot of rate cuts. And now all of a sudden, that doesn't look like it's going to happen. And obviously, currencies and interest rates are very closely aligned. So if you look what I gave you the 12/31 number, so obviously, the dollar has gotten a bit stronger. I gave you the numbers in terms of the total non-U.S. dollars. So you can kind of do some math, it would still be in that range. Obviously, it would be a little bit of a headwind just given that the dollar has gotten stronger since 12/31. But I think you can take our Q1 guidance and sort of fold that in and think about the normal seasonality that you have and come up with an answer, but it would still be in that range, though.

Just to clarify, so it's still in that 6% to 8% range if we put it in the USD or are you saying it's still kind of mid-single digit, but maybe something like a point lower?

Well, I'd say if you're using the spot rate as of today, the simple math would suggest that it is. So if you look at -- just take the midpoint of the guide. If I just use that and what the impact of the dollar has been, it's still in that range. Now you had asked it could be [indiscernible]. Obviously, the dollar was at [ 101 ] back at 12/31, it was at [ 106 ] in November. It's bouncing all over the place. Obviously, if it were to move, you can do the math. 5 or 6 or points lower, you could potentially get on the other side of that. So again, it's just -- I'd be end up giving you guys a massive range that wouldn't be helpful. So what I'd rather do is just give you guys the tools that you can do it yourself and look at really get an understanding of the core business underneath that. [ How was that ] that I think is much more important to understand.

The next question comes from Madeline Brooks of Bank of America.

Just one on security. The outside of Guardicore, if [ you ] want to touch on this year, the rest of the [ Zero Trust ] portfolio trends that you've seen and maybe if you're feeling any additional competitive pressure now that the market has really expanded there? And then I have one follow-up.

Yes. In web app firewall we've been a market leader there for 10 years since we started that marketplace with Web App Firewall as a Service. And after 10 years, you do get competition, but we're still the market leader by a good margin, and that's a good growth business for us. We've added a lot of capabilities on top. Bot management and more recently, account protector, client-side protection so that customers of commerce sites can stay safe by going to the site, you're going to need that now for compliance. There's brand protector. So that's identifying the fishing sites and keeping them from stealing user information. Of course, we've been doing a denial of service protection for a long, long time now, market leadership position there. And then you have on the enterprise side, of course, Guardicore we talked about doing very well. And I'm really excited about API security. I think over the longer term, that becomes as big a marketplace and just as important as web app firewall has become. And our goal there is to become the market leader and already in the go-to-market motion, there's a strong synergy between Web App Firewall and API security. We built a very easy way to do a proof of concept for our web app firewall customers, and that's where we're getting a lot of early traction. Also, we've integrated with a lot of the load balancers and other firewalls out there so that we can sign on new customers who are not using Akamai CDN or a web app firewall. So I think there's a variety of areas in security that are working very well for us.

And then just one quick one on [ Compute too ]. I think if you think about earnings that have happened so far, especially with hyperscalers like AWS, Microsoft, Meta, we've kind of heard of the theme of the optimization inflection in terms of cloud computing meaning maybe this year, we're going to see a little bit more investment in new workloads. I'm just wondering if you've heard of any of those trends among your customers who are thinking about compute for the first time or maybe if you're seeing increased appetite for compute for this coming year versus 2023?

Yes, it Compute is an enormous marketplace and growing rapidly, and there's always new applications that are being created and not just migrating from a data center into the cloud, but just brand-new applications. And so that's where we're seeing a lot of traction also in some cases, lift and shift out of a data center or out of a hyperscaler. But it's just an enormous marketplace and a great place for us to operate. And even those that are optimizing that sort of, I guess, not such a great thing for the hyperscalers. But we're part of that trend. It's great for us because we can help customers reduce cloud spend. And we've gotten very good feedback from our early adopters of Akamai Connect to cloud that they're saving a lot of money. So the trend to optimization is a positive thing for Akamai.

The next question comes from Rishi Jaluria of RBC.

I wanted to drill on to maybe going back to Gecko. I guess, number one, can you talk a little bit about edge inferencing and what those use cases look like? It's one of those things that we hear a lot of talk about in theory, but maybe in practicality as you're talking to your customers and having those conversations, what can that look like? And what positions Gecko uniquely for that? And then maybe financially, to the extent, I know it's still early, are you assuming real Gecko contribution on the compute line in your guidance for the year? or is that something that as it gets traction, could lead to more upside beyond what your model?

Great. So let me start with edge inferencing. And so some of the examples I gave, that's exactly what's happening for commerce sites in figuring out in real time, what content you're actually going to give to the user that's coming to the site, ad targeting, what ad do they get? Anything that involves personalization on the security side, a ton of inferencing is used to analyze real-time data. For example, even our own bond management solution, is that entity that's coming to the site? Is it a bot or is it a human? And even if it's a human and they have the right credentials is it the right human? And you use AI and inferencing for that, and you got to do it really fast. You can't -- before to send it back to the centralized data center because you got a massive number of people that you've got to process in real time, especially if you're doing some kind of live events. And so being at the edge matters because you can be scalable, you can handle it locally, you get great performance. You can make it be real time. And Akamai's unique value proposition with Gecko is that we're going to be able to now support this not in a few cities, but in 100 cities by the end of this year. So anything you can put in a VM, virtual machine, which is most things you're going to be able to do that in 100 cities. And then ultimately, in hundreds of cities because we can put this in general Akamai Edge [ pops ]. And then next will be containers, which is pretty much the rest of what you do in cloud computing. And then to be able to spin it all up automatically. It's a whole new concept for compute that I think is very powerful. And there's a high overlap of wanting to do that with inferencing engines, where you're trying to do something intelligent based on that end user or that entity that's interacting with the application. Now in terms of Gecko, we're just now in the early stages of getting it deployed. We're in 9 cities, we'll get the tenth new city up in another month or so. By the end of the year, we'll be in a total of 100 cities supporting compute. So not a lot of revenue is factored into the guidance based on Gecko for this year. That would come more next year. So this year's revenue guidance is based on the original 25 core compute regions that we've set up by the end of last year. Now we will deploy this, of course, just as fast as we can and as customers want to adopt it. And hopefully, we have the situation where we want to build out more, get more compute capacity because there's so much demand for compute on Akamai. Ed, do you have any color you want to add around the guidance there?

No, I think you captured it right, Tom. We don't -- we're not really anticipating anything. I mean one thing we are doing this year is we are making changes to our comp plan with our reps so that they all have to sell compute this year. So you could see things, reps get very creative, I learned in sales that comp drives behavior. So by leaning in here and making it something that all reps have to do, we should see a lot more use cases, a lot more opportunities, et cetera. So there's always a chance that we could be surprised here with the creativity of our field, bringing us opportunities, but we did not factor in anything material as it relates to Gecko.

The next question comes from Michael Elias of TD Cowen.

Just first on Gecko. Presumably, the Pops that you have, they are already supporting security and delivery workloads. Just from an architecture perspective, can you help us think about what expanding the compute platform into these Pops means? Is it just additional colocation deployments and on the CapEx side, networking gear and servers. Any color that you could give there in terms of the mechanics of what the expansion would look like? And then second, Ed, last year you were talking about elongation of enterprise sales cycles. Just curious what you're seeing in terms of the buying behavior of your customer base? Any notable call out there?

All right. So I'll take the first one. With Gecko, that is generally speaking an existing Akamai Edge pops. And in particular, they tend to be the larger ones where we already have a lot of equipment that's already connected into our backbone, and what we'd be doing is adding additional servers. And for compute, it would be a beefier server and additional [ colo ] for those servers. But all the other infrastructure is generally already there, and it's already connected in and we already have delivery and security operating there. So it does become a very efficient way for us to deploy Gecko. And Ed, do you want to take the second one there?

Sure. Yes. So I think the trend, obviously, acquiring new customers is always challenging in an environment like this. The one probably exception to that is in the security space. That tends to be something that obviously, now with the requirements with the SEC reporting and I added a disclosure with CISOs being now potentially criminally charged for breaches and things like that. Audit Committee is spending more and more time on cybersecurity as a topic. That tends to be a budget that, one, you don't typically cut and, two, you're generally adding to. But yes, new customers are challenging. I do think this kind of environment helps us what we were just talking about in the last few questions, around optimizing cloud spend. Certainly, if you've seen what we've done, we've [ seen ] a tremendous amount of money. So I think that can also help us in this particular environment. But definitely, new customer acquisition is a bit more challenging, but we're still doing pretty well. Obviously, the environment can change. But it hasn't been a major factor for us yet.

The next question comes from Ray McDonough of Guggenheim Securities.

Maybe, Tom, just a follow-up to a prior question. As we think about Gecko and you mentioned that your edge sites right now don't have full stack compute. But how much work is done to be to converge what you already have at your edge sites and what you've done in terms of building out the nodes capabilities? Should we expect there to be a common software stack across both edge and centralized sites? And if so, is the plan to have that in place by year-end?

Yes. Great question. So what we're doing now is, as I mentioned in the last question, deploying more hardware in existing edge region and generally the larger edge regions. We already have network there. We already have delivery security located there. So there's some additional color and servers. And yes, the goal is to put it all on one common software stack. Now initially, we have the [ node ] stack is moving into these edge pops for Gecko. But as we once we get the support for virtual machines and containers, then next, we want to add the software stack that we have for delivery and for security and for function as a service that automatically, for example, spins up JavaScript apps and milliseconds based on end-user demand, we want all of that to be operating on containers and VMs. So that you don't have to think ahead of time about how many VMs do you want in each of these hundreds of cities. It just happens based on end user demand. You automatically get new ones spun up, load balancing, fail over, really a very compelling concept. And that doesn't exist in the cloud marketplace today. That is the vision -- I think you really nailed it when you talked about the common software stack because only Akamai has that full edge platform today, that software stack around delivery and security that will be now including compute.

And maybe as a quick follow-up, as we think about the expansion of the Linode footprint last year, can you help us understand as much as you can, how much of the space currently built out was for internal use purposes to help with that third-party cloud savings versus space that's online now that's revenue generating? And I know we've talked about this in the past, but any color around what we should expect from a customer utilization perspective that might be embedded in your guidance as we move through year-end? That would be helpful.

Yes, sure. So the majority of the growth in the compute guidance is going to come from enterprise compute. So the stuff that we built out for the last year. So if you kind of go back and look at the math in terms of -- we've kind of said roughly speaking, $1 of CapEx is $1 of revenue. You can kind of look at what we're doing for compute build-out now what we did last year. We said we got about $100 million roughly for our -- all in for our internal use. So that leaves a pretty significant amount left for customer demand. Now obviously, the way people buy today, they pick a location, et cetera. So it's not going to be exactly a dollar for dollar right now, but it's a general rule of thumb. So I would say the majority of what we built out is for customer usage.

The next question comes from Tim Horan of Oppenheimer.

Following up on Ray's question. So I'm assuming the goal here is to get to one single platform where customers can access the full range of services relatively easily on, I guess, one on ramp up. When do you think you'll get there? And secondly, the Gecko product, it sounds like is this completely serverless? And is it a development platform also?

Yes. So I think one platform really in terms of being able to do everything together and all the same software so that we have our Edge software running with the Linode software to spin up VMs and containers that's not until 2025. We are, first, combining the infrastructure and of course, customers can buy the services as a package. We have common reporting now in many cases. But in terms of doing all the automatic spinning up and truly serverless used for VMs and containers, think 2025 for that. And let's see. So -- and what was the other question you had?

So the new product, Gecko, it is primarily a serverless product, it sounds like. And do you have all the support there for developers to completely run their applications on this new platform?

Yes. And it's -- it depends how you define serverless, initially with Gecko, you would operate it the same way you would Linode. You decide how many VMs and containers you want in the various cities. And it is very developer-friendly, works just like Linode. So if you're familiar there, that would now work in -- well, at the end of the year, 100 cities for your virtual machines. Now if you define serverless to be -- which doesn't exist today out in the marketplace for VMs and containers, they just spin up automatically like we do today for Function as a Service and for JavaScript, that's what comes 2025.

If you don't mind me asking me, it sounds a lot like what Cloudflare is doing, but you're saying it doesn't exist today. Can you maybe talk about a little bit what's different, what you're doing?

Yes, that's a great question. They don't support VMs or containers at all, never mind serverless or anything else, just -- they don't have support for that. They don't do this full stack cloud computing.

The next question comes from Alex Henderson of Needham.

Great. So it seems pretty clear that Guardicore is a critical piece of your security growth. And obviously is perturbing the overall growth rate. I was hoping you could give us some sense of what the security product lines, excluding Guardicore look like in terms of their growth rates. Any sizing of that growth would be even a ballpark would be quite helpful. And then second, I was hoping you could talk a little bit about your -- you mentioned inferencing, but I think it came in kind of as an afterthought as opposed to the primary focus. Can you talk about your involvement in AI inferencing at the edge and to what extent that requires either the 2025 kind of structure or what needs you have there and whether you're putting GPUs out of the edge in order to facilitate that?

Ed, do you want to go with the first one, and I'll take the second one?

Sure. Why don't I go to the first one. On the spirit of it being a year-end call, I'll break out these numbers at a high level for you, but won't be doing it every quarter. So if I look at the -- what we used to -- what we call the [ Appen ] API security bucket, that's our largest bucket, that includes bot management, our fraud products, our web app firewall, our new API security product. That's actually in Q4 growing over 20%. So that's been incredible. Guardicore itself, if I normalize for the onetime software that we did last Q4 is growing at about 6%. And infrastructure and services are growing sub-10%.

Just to be clear, is this the full year growth rate? Or is this the fourth quarter growth rate?

This is the fourth quarter growth rate. The full year I don't have [indiscernible].

That's efficient. I just needed to know what it was.

Okay. Yes, in terms of the question around inferencing and AI and so forth, yes, we're building full stack compute to have great performance at a lower price point and have that available in hundreds of cities. And one of the many things that you would do with that is AI inferencing and that's not an afterthought. In fact, we've been using AI in our products for, well, 10 years, bot management, for example, runs on Akamai connected cloud. It's one of many things that run on it. So not an afterthought. There is an enormous amount of buzz now about AI. And I think a lot of that is justified. And I think there's a lot more compute going to be consumed because of AI. And it is a strong use case among our customers that are using Akamai connected cloud. That said, it's not all AI. In fact, our biggest customers are doing media workflow, doing live transcoding. And that's not using AI. So I think AI is an important use case, one of several use cases. Now in particular, you asked [ '24 versus '25 ] it's being done already on our platforms. There's no need to wait until the end of the year unless you want to do it in 100 cities, then that comes at the end of the year. No need to wait to [ '25 ] when the instances are spun up automatically instead of by design ahead of time the way compute works today. So you talked about GPUs. Akamai has GPUs deployed, we're deploying more. We've used them in the past for graphics. And going forward, probably use them for Gen AI uses. We're not really deploying them right now in the edge Pops. And that's not -- it just because you don't need to. It's not cost-effective. And the edge pops, you're going to be doing the inferencing. And for the inferencing, you can use GPUs, but we're also using CPUs. And right now, we get a better ROI on the CPUs. So I guess there's a lot of confusion there as well. Now GPUs are critical for doing training for especially large language models and that's going to be done in the core and we're not supporting that as a key use case today. We could, in theory, right, we have all the technology to do it, but that's not where we're focused in terms of getting the best ROI for our platform. And for that matter, most of the work with these models, most of the compute is done when you're using them for the inferencing. You do the training, and then you spend -- just so it learns you get it ready to go, and then you operate it. And it's the operation where most the vast majority of the cycles are and that can be done on CPUs. And in many cases, the cases I mentioned, for a personalization for security, for data analytics, that's done on the edge more as good reason to be done there using CPU-based hardware.

The next question comes from Jonathan Ho of William Blair.

Just one question from me. How important is the global load balancing capability? And what does that maybe mean for your ability to either attract more customers or to drive revenue from that product?

Yes, that's very helpful because it makes it much more scalable. You have failover, so much more reliable. And I think it's the basic capability, of course, we've had forever, it seems in delivery of security, and now that's available for compute. So I think that's important and that greatly increases the market we can go after for compute.

And operator, we have time for one last question.

Our last question will come from Rudy Kessinger of D.A. Davidson.

Ed if my math is correct, even if I exclude the $100 million in CapEx and compute last year, intended for moving over internal workloads between last year and this year, it looks to be about $400 million in compute CapEx. And going back to that kind of $1 in CapEx equals $1 of revenue capacity, $400 million in CapEx, roughly $200 million of compute growth '24 versus '22. Do you feel like you guys are maybe over building at all? Or what gives you the confidence, I guess, in the pipeline and the ramping usage to spend so much on another round of build-out this year when we're not yet seeing growth accelerate, right? You're guiding to 20% growth next year that's flat with Q4.

Yes. So I'll just -- let me address that part first. So I would say if you look at the underlying components of what's growing, it's actually that enterprise compute opportunity that's growing very, very, very fast, like those numbers, the percentages would be kind of foolish to break out because they're going off of small numbers and adding to them very big numbers. Now also part of our strategy is to be competitive and have big core centers in many cities, and that does require a larger build out. So there's a lot of capacity that we have to sell. And then also, we're seeing demand in certain cities, you have to build out more capacity where you're getting demand. And then the Gecko sites that we're building out, it's not a significant -- I mean it's a decent amount of capital. But I think that is another big key differentiator for us. And as Tom mentioned, we think there's a big opportunity there. So I know a lot of people have been questioning us being able to take on large workloads, et cetera. We clearly have a lot of capacity out there. As I talked about earlier, we've made the change with our compensation plans where our reps now have to sell compute. So we're going to see a lot more of that. We've done a tremendous amount with the platform in terms of adding functionality. We built out the platform, connected it to our backbone, and we have a lot of new compute partners. The platform is ready to be sold so we're pretty optimistic about it. And I think we're building in a pretty responsible manner. As I talked about, our CapEx is relatively modest for this business right now. So I think we're in pretty good shape. And with that, that will end today's call. I want to thank everyone for joining, and have a great evening.

The conference has now concluded. Thank you for attending today's presentation, and you may now disconnect.